This revised R21 proposal in response to PAR-11-216 (Early Phase Clinical Trials in Imaging and Image- Guided Interventions) will evaluate a new imaging and image-guided intervention for resection of breast cancers which are not palpable to (i) eliminate wire localization - a moderately effective, somewhat costly and certainly inconvenient procedure for patients - and concomitantly to (ii) reduce positive margin rates and subsequent re-excision surgeries - a significant but unnecessary burden not only on patients but also on the cost of healthcare. Specifically, we have combined pre-operative supine breast MR with intraoperative optically scanned fiducials and non-rigid surface-to-volume image registration to provide the surgeon with real-time visualization and tracking of the surgical fieldin the operating room (OR) based on the deformed and coregistered (to the surgical position) preoperative supine MR images for cancer localization and surgical (complete) resection. We have also completed Stage I of the proposed clinical study during which we evaluated how closely surface-projections of tumor position determined from the coregistered, optical-scan- adjusted MR image volume corresponded to the location found through the surgeon's manual measurements of tumor position based on palpation at time of surgery (in Stage 1, we purposely enrolled women with palpable breast cancers to provide independent comparison data for evaluating our coregistration procedures in the OR). We met our IRB-approved acceptance criterion of all intraoperative image-guidance measurements of tumor location being within 1 cm of the palpable tumor edge in 5 consecutive OR cases (our last 7 cases had a mean difference of 0.37 cm + 0.19 cm), and we expect an accuracy of < 0.5 cm will be achieved in the proposed Stage II study which is well within the 1 cm margin of tissue that surgeons routinely add to the intended resection volume when wire localization is used. We now propose a single institution Stage II study of supine MR imaging and image-guided resection of non-palpable breast cancers with the primary objective to determine if the supine image-guided breast conserving surgery (BCS) in non-palpable breast cancers results in lower positive margin and re-excision rates relative to surgeries in patients undergoing standard-of-care wire localization with presurgical planning based on prone MR image volumes that are not coregistered with the operative field, in a randomized enrollment into these two groups. The positive margin rates associated with contemporary BCS are discouraging (~25%-50%8-11), and currently represent significant strains on patients and the cost of breast healthcare. Meta-analysis of the impact of surgical margins on local recurrence also confirms that margin status has a positive prognostic effect 7a. Thus, we are in a significant position to explore a potentially practice-changing approach to BCS for breast cancers that are not palpable.